[dpdk-dev] [PATCH v2 4/6] bond: free mbufs if transmission fails in bonding tx_burst functions
Doherty, Declan
declan.doherty at intel.com
Tue Sep 2 11:22:31 CEST 2014
> -----Original Message-----
> From: Doherty, Declan
> Sent: Monday, September 1, 2014 9:31 AM
> To: dev at dpdk.org
> Cc: thomas.monjalon at 6wind.com; rsanford at akamai.com; Doherty, Declan
> Subject: [PATCH v2 4/6] bond: free mbufs if transmission fails in bonding tx_burst
> functions
>
> Fixing a number of corner cases that if transmission failed on slave devices then
> this
> could lead to leaked mbufs
>
> V2 addresses behaviouraly issues in the first version and packets are no longer
> freed in
> the bonding layer, except in the case of broadcast mode where in failures happen
> on
> more than a single slave then mbufs will be freed in all slaves except the one
> where
> the least errors occured. Also contains new unit tests to test the transmission
> failure case in slaves for
> round-robin, balance, and broadcast modes.
>
>
> Signed-off-by: Declan Doherty <declan.doherty at intel.com>
> ---
.....
> --
> 1.7.0.7
I've just noticed that I inserted an extra character into the patch when I was annotating this
patch which will cause this a fail compilation.
> + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE -s
> + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT);
should have been
> + TEST_BCAST_SLAVE_TX_FAIL_BURST_SIZE -
> + TEST_BCAST_SLAVE_TX_FAIL_MIN_PACKETS_COUNT);
Thomas, do you want me to submit a V3 of this?
Regards
Declan
More information about the dev
mailing list